The overarching goal of this research project is to revise and enhance a multimedia computer-based multiple risk factor intervention for cancer prevention, using the internet to reach a general population. Two innovative, individualized, easily disseminated, low-cost, and interactive interventions, both for multiple behaviors (sun protection and exercise adoption), will be developed and evaluated in comparison to a control condition. The first intervention involves adapting our effective multimedia expert system interventions to the internet environment. The second intervention builds on the first with the additional inclusion of a Relational Agent, a recently developed computer-based approach to establishing a personal relationship typically missing on internet sites. Both systems will employ the same theoretical model of behavior change, the Transtheoretical Model, as the deep knowledge, and all systems will employ empirically based decision rules. The design is a 3 Group (Control, Internet, Internet plus Relational Agent) x 3 Occasions (0, 12, 24 Months) with intervention occurring during the first 12 months. A representative national sample of 1639 individuals at risk for both behaviors will be recruited. The primary aims are: (1) To develop and assess the effectiveness of a tailored internet intervention on a national sample;(2) To develop and assess the effectiveness of the internet intervention enhanced by a relational agent;and (3) To determine if the intervention with the relational agent can outperform the regular tailored internet intervention. The secondary aims are: (1) To determine if the two interventions are differentially effective with each behavior and with different subgroups, and (2) To determine if the relational agent intervention is utilized more often for increasing exercise than for sun protection. This study targets two major risk factors for cancer;is designed to treat the behaviors on a population basis, using proactive recruitment strategies;intervenes on multiple behaviors simultaneously, thereby producing greater impacts for cancer prevention;utilizes one of the most promising approaches to low cost population based interventions for health-related behavior change, namely the internet;and develops and tests a promising new approach to increasing the utilization and effectiveness of internet-based interventions, relational agents.